Xinlian Power completes nearly RMB 2 billion Series C financing at a post-money valuation exceeding USD 2 billion, led by China Mobile Chain-Length Fund

Jiemian News has learned exclusively that Xinlian Power (hereinafter referred to as "the Company") has announced the completion of a Series C funding round of nearly RMB 2 billion. The round was led by China Mobile Chain-Length Fund, with participation from investors including a state-owned power grid central enterprise, Inspur Information, Yuanjing Capital, Sinopec Hydrogen Energy Fund, NAURA Novachip Capital, Jiangfeng Electronics, Changzhou Xingyu, Cathay Capital, Puchao Capital, CCB Equity, CITIC Jinshi, CITIC Securities, and Galaxy Yuanhui.

The change in the investor lineup reveals this intention. Three years ago, when Xinlian Power was founded, its founding shareholders were exclusively leading players in the automotive and new energy industries; in this round, the shareholder lineup has expanded to the communications, computing power, and power grid sectors, with the entry of several industrial capital players in the AI infrastructure field particularly notable.

Founded in 2023, Xinlian Power is a one-stop system solutions provider for compound semiconductor manufacturing and module packaging, focusing on the design, R&D, production, and sales of compound semiconductors such as silicon carbide (SiC) and gallium nitride (GaN) and module packaging. According to a 2025 Yole report, the Company's silicon carbide business ranks first in China and among the top five globally.

New energy vehicles and AI data centers appear far apart, but in high-voltage power systems, silicon carbide is an important device for both, giving Xinlian Power a foundation for capability migration. The difficulties are also obvious: beyond launching suitable products as quickly as possible, the application scenarios, technical architecture, and industry players have all changed. This also explains why the Company needed to gather a group of new partners in this funding round.

Nearly RMB 2 billion is a considerable large-scale financing. It is understood that this round was not broadly publicized and was communicated only on a targeted basis with some investors, but the news gradually spread among industry institutions, ultimately attracting more than RMB 4 billion in investment intentions. After completion of this round, Xinlian Power's valuation exceeds USD 2 billion.

On the surface, this is yet another investment subscription in which institutions competed for allocation. But "being unable to secure an allocation" is not enough to explain this round.

The investor list is in fact an "ally" list: most participants have clear directions for business synergy, and some cooperation has already entered the stages of joint chip research, product sampling, industrial investment, or project incubation. Equity investment is not the starting point of cooperation, but rather more like the result of deepened cooperation between the two sides.

This is not the first time Xinlian Power has "gathered people around a cause." When it was founded in 2023, Xinlian Power brought together automotive and new energy industry players such as SAIC, XPeng, Bosch, CATL, and Sungrow as founding shareholders. For details, please see "A RMB 40 Billion Chip Giant Builds a 'Circle of Friends' in Shaoxing."

In an upcycle, opening a second curve

The AI boom has directly brought a surge in computing power, and the number and scale of data centers have soared accordingly. ABI Research predicts that by 2030, the number of data centers invested in globally will exceed 8,400.

Pushing the question further down, what ultimately constrains computing power? According to the "AI five-layer cake" proposed by Jensen Huang, the bottom layer is not models, nor chips, but energy. Sam Altman has also repeatedly regarded energy as a key condition for AI's continued expansion.

America's constraint is "data centers exist, but there is no power to connect to." The expansion speed of large model companies and cloud vendors far exceeds the pace of power grid upgrades, and data centers have begun to "chase power."

China has a more unified power grid system and stronger infrastructure construction capabilities. The real problem is: the power is there, so how does computing power follow it?

China's computing power demand is mainly concentrated in the east, while energy resources such as wind and solar are more distributed in the west. In the past it was "west-to-east power transmission"; now "east-to-west computing" has further emerged. The key behind this is not just increasing power generation, but computing-power-electricity coordination—planning computing hubs, power sources, power grids, and energy storage together.

In 2026, "computing-power-electricity coordination" was written into the government work report for the first time and incorporated into the outline of the 15th Five-Year Plan. The coordination of computing power and electricity has begun to rise from market discussion to a national-level infrastructure project.

The above constitutes the premise for Xinlian Power's "gathering people around a cause" at the "people alignment" level.

Another change is taking place inside data centers. Data centers are moving from the traditional tens of megawatts toward hundreds of megawatts or even gigawatts. According to reports, Deepseek plans to build a 1GW-class data center in Ulanqab.

As the scale of data centers expands, the power of a single rack rises linearly, and traditional power supply architectures begin to face a series of problems: losses from multi-stage power conversion, the need for more copper materials to carry large currents, power and cooling equipment crowding out machine room space, and the higher requirements that rapid GPU load fluctuations place on grid stability, among others.

At present, the main technical routes are being advanced in layers: the faster-to-deploy one is 800V high-voltage direct current power supply, namely HVDC; the medium- and long-term direction is the solid-state transformer (SST), which directly converts medium-voltage alternating current into high-voltage direct current. The two transform different levels of the power supply path, and both aim to shorten the power supply chain, reduce losses, support higher-power computing clusters within limited space, and achieve cost reduction and efficiency improvement.

This change brings new opportunities for power semiconductors. Compared with traditional silicon-based devices, silicon carbide can withstand higher voltage, temperature, and power, while gallium nitride is suitable for high-frequency, high-power-density conversion scenarios. These two types of compound semiconductors precisely enter links such as SST, HVDC, UPS, and server power supplies, respectively.

Among them, silicon carbide power semiconductors are regarded as one of the links with the greatest value and the highest barriers in solid-state transformers. According to data published by Infineon and solid-state transformer company DG Matrix, compared with traditional transformers, the volume of solid-state transformers can be reduced to 1/14 of the former, and the weight to about 1/40.

A few months ago, Nvidia had just brought seven SST manufacturers, including DG Matrix, into the Nvidia MGX ecosystem as solid-state transformer suppliers for its 800V DC power supply architecture. Changes in data center power supply architecture have already moved from concept to products and supply chains.

One point is worth noting: this second curve is a new direction that Silicon Mobility proactively opened up during its upcycle. In the first half of 2026, Silicon Mobility achieved revenue of 1.182 billion yuan, a year-on-year increase of about 106%; full-year revenue is expected to exceed 3 billion yuan, and the automotive business is still in a period of rapid growth.

Silicon Mobility chose to "fix the roof while the sun is shining" during its upcycle, proactively extending its automotive silicon carbide manufacturing capabilities into AI power supply. The judgment is mainly based on three conditions: First, AI infrastructure itself is a sufficiently large incremental market. According to Yuan Feng's judgment, AI infrastructure may be one of the few ten-trillion-level markets after real estate and automobiles, with power-related investment potentially accounting for about 10%.

Macro-level forecasts are one aspect. According to media reports previously citing people familiar with the matter, China plans to invest about 2 trillion yuan over the next five years to build a nationwide computing power network, with state-owned operators such as China Mobile and China Telecom likely to undertake the main operational work.

AI infrastructure is undergoing drastic technological change and supply chain restructuring, providing an opening for new entrants. The opportunity once brought by new energy vehicles came from electrification and intelligentization driving simultaneous changes in powertrains, electronic architectures, and supply chains. Data centers moving from traditional low-voltage power supply to HVDC and SST will likewise change product definitions and supplier relationships.

Similar to the opportunity in new energy vehicles, the previous wave of opportunity came from the "four modernizations" driving simultaneous changes in powertrains, electronic architectures, and supply chains. Data centers moving from traditional low-voltage power supply to HVDC and SST will likewise change product definitions and supplier relationships. What Silicon Mobility hopes to seize is precisely the window of supply chain rearrangement.

Three years ago, Silicon Mobility built capabilities around new energy vehicles covering silicon carbide chip design, wafer manufacturing, module packaging, and large-scale mass production. Automobiles mainly use 750V and 1200V products, while AI power supply requires extending devices to higher voltages, but the underlying foundation is still centered on power semiconductors.

Therefore, what industry players truly buy into is not a particular automotive chip from Silicon Mobility, but the company's comprehensive capabilities in designing, manufacturing, and mass-producing high-performance power devices.

What can be transferred is chip design, process manufacturing, module packaging, reliability verification, and large-scale mass production capabilities; what cannot be directly replicated is customer relationships, application experience, and product definition.

The automotive business has proven that Silicon Mobility can make chips and complete large-scale delivery; the AI business requires it to re-understand the power supply architecture of data centers, re-recognize power grids, cloud vendors, server companies, and power supply enterprises, and jointly define products with them.

Replicating "gathering people around a cause"

A new round of "making friends" began with a top-down review. Silicon Mobility first reviewed the key links in the industrial chain, and its new partners in "computing-power coordination" cover computing power, communications, power grids, new energy, and more.

The most upstream is the power grid, the starting point of power supply; China Mobile is both an operator and a builder of cloud and computing infrastructure; Yuanjing Capital can connect Alibaba Cloud and its AI infrastructure ecosystem resources, and Alibaba is also one of the earlier domestic enterprises to study AI data center power supply architecture from the system level; Inspur Information is a server and AI infrastructure giant; NAURA and Jiangfeng Electronics are located in the equipment and materials links of the semiconductor industry chain.

This round of financing was not without financial investment institutions, but overall it showed a very distinct industrial character. Even for financial institutions, what Silicon Mobility values is not just capital, but the industrial resources they can connect to behind them. For example, behind Cathay Capital is a link to France and Europe's energy industry, a potential partner for future business globalization; CCB Equity and CITIC Securities correspond to resources at the financing and capital market levels.

At this point, the key nodes in the AI power supply chain are basically covered. It is understood that Silicon Mobility has carried out substantive business cooperation with some of the above companies in areas such as joint chip research, industrial investment, and project incubation. Silicon Mobility currently covers silicon carbide products across the full voltage range from 650V to 3300V in the AIDC field, among which the 3300V SiC MOSFET has been sent to core customers for sample verification, and products at 6.5kV and above are in the R&D stage.

These cooperations were not discussed only after financing, but rather business recognition came first and financing followed. Before taking equity stakes, Silicon Mobility had already conducted business communication, technical exchanges, or product sample delivery with them. Therefore, the 2 billion yuan financing is not the starting point but the result, merely a seal of approval that adds weight to the cooperative relationships.

Using CVC as a path, or establishing joint ventures to achieve a certain business goal together, is certainly nothing new, but these instances of "gathering people around a cause" by Silicon Mobility are also hard to call typical CVC; they carry a distinct industrial imprint and personal experience.

Yuan Feng, the person driving Silicon Mobility's financing, has a distinct CVC background. He joined Silicon Integrated in 2023 and was previously CEO of GAC Capital, long focusing on the vehicle business and solving issues such as automotive chip shortages, supply chain stability, and new technology layout through investment.

Three years ago, the "cause" Silicon Mobility wanted to pursue was to promote domestic silicon carbide into new energy vehicles. When the company was founded in 2023, it brought together automotive and new energy industry players such as SAIC, XPeng, Bosch, CATL, and Sungrow.

At the time, China's new energy vehicle market was growing rapidly, automakers had just gone through a chip shortage, 800V high-voltage platforms were beginning to become widespread, yet the silicon carbide market was still mainly controlled by overseas companies. Relying only on traditional procurement relationships made it difficult to solve product iteration and supply security issues. So potential customers entered the product definition and R&D stages early, the Silicon Integrated system handled chip manufacturing and module packaging, and investment was then used to further lock in the relationship between the two sides.

The hybrid silicon carbide module jointly developed by XPeng and Silicon Integrated is one example of this model. The two sides combined silicon-based IGBTs with SiC MOSFETs to find a new balance between efficiency and cost, and the product was subsequently used in the XPeng X9. This was no longer a simple substitution based on overseas standard products, but rather an automaker and a chip company jointly defining a new product around a specific vehicle model.

Three years later, the company expects revenue to exceed 3 billion yuan in 2026. According to NE Times statistics, in the first half of this year, Silicon Mobility's SiC power modules ranked first in China with a 16.9% domestic installation share.

Now Silicon Mobility is trying to replicate this experience in AI infrastructure. The reason for establishing it again is the logic of the AI-adapted supply chain industry—each additional user and each inference consumes more tokens and adds more GPUs, optical modules, data centers, and power. The expansion of AI companies no longer depends only on users and models, but also on whether they can obtain enough chips, energy, and infrastructure.

The United States has already provided a sample. The main players in computing centers are big companies; AI giants such as Nvidia, Microsoft, Amazon, and OpenAI are all actively making upstream layouts along the AI supply chain, using investment to lock in chips, advanced packaging, data centers, and power resources.

The same principle applies in the Chinese market, and the player profiles are even more diverse. Power grids, operators, cloud vendors, servers, and chip companies are all involved, each possessing part of the capabilities, but no single company can complete the entire construction from power generation to GPU power supply on its own. This creates the possibility of alliances and combinations like those in new energy vehicles.

The so-called "replication" is repeating a method of organizing industry: first find a major issue in the midst of technological change, then organize the application side, system side, chip side, and manufacturing side to participate together from the product definition stage.

So Silicon Mobility is far from being merely an invested company; it is a connector within the Silicon Integrated system and itself serves as a bond for industrial collaboration. This method has a very distinct CVC flavor, but it is not entirely equivalent to traditional supply chain investment.

In the past, supply chain investment usually meant that a company discovered it lacked a certain product or capability and then invested in a corresponding supplier. What Yuan Feng calls "gathering people around a cause" is, when a new industrial chain has not yet fully taken shape, first determining the problem that needs to be solved together, then organizing companies with different capabilities to jointly define products, incubate projects, and promote the implementation of technology.

From this perspective, Silicon Mobility is not just an invested company. As a product company within the Silicon Integrated system that connects major customers, it itself also serves as a bond for industrial collaboration.

Equity relationships can increase trust and cooperation efficiency, allowing Silicon Mobility to get closer to customers earlier than a company that enters the supply chain only through quotations, and to participate in product definition earlier. But capital cannot replace products, and industrial shareholders cannot be directly equated with orders. The final result still depends on chip performance, reliability, cost, and delivery capability.

The 2 billion yuan financing and a list of industrial shareholders are only the beginning. The real result still depends on whether it can move from capital alliances to joint product development and ultimately appear in AI data centers one after another.

This article comes from the WeChat public account "ChinaVenture", author: Cao Weiyu, published by 36Kr with authorization.